Been a lurker for a relatively short time, took the survey.
I had some concerns over the extra credit questions and one thing in particular that prompted me to respond. I agree it seems there was meant to be no right answer to a couple of the questions, and the babies in the hospital was at least a clear statistical problem. I also had an admittedly whimsical objection to the lack of details on one question, thanks to the level of specificity seen in riddles, puzzles, and so on here, and maybe due to a programmer background thinking of pointers and assignments. The first CFAR question should have specified to start something like there are three people in a room. Then it’s clear there’s not a person looking at himself in a mirror, or more than three people with some having the same first name, or creatively a statue or painting, with one human looking at the artwork which is facing another human. (Would be a good one with Lisa implying the Mona Lisa, but connotations of names shouldn’t be relevant)
However, what I was really curious about was the redwood question. Surprisingly, I knew the answer pretty much exactly and later noticed the complaints about using feet as a unit which was rather strange. I remembered this exact question was from a recent study in a psychology journal on cognitive biases.
It seems likely the question was written with knowledge of the study in mind, to test for anchoring bias, considering many other things could have been chosen other than the height of redwood trees in feet. I really hope the intention was not to take these results to say, “LW readers conclusively have less anchoring bias than people in the psychology study.” As the maximum number that could possibly be generated by the RNG (999) iirc is lower than an anchoring number given in the study (1000), presumably few survey participants if any would answer the tree height is greater than that. Due to the RNG anchor likely being much closer to the true value anchoring bias will be less than in the psychology study by nature of how the question was presented.
Been a lurker for a relatively short time, took the survey.
I had some concerns over the extra credit questions and one thing in particular that prompted me to respond. I agree it seems there was meant to be no right answer to a couple of the questions, and the babies in the hospital was at least a clear statistical problem. I also had an admittedly whimsical objection to the lack of details on one question, thanks to the level of specificity seen in riddles, puzzles, and so on here, and maybe due to a programmer background thinking of pointers and assignments. The first CFAR question should have specified to start something like there are three people in a room. Then it’s clear there’s not a person looking at himself in a mirror, or more than three people with some having the same first name, or creatively a statue or painting, with one human looking at the artwork which is facing another human. (Would be a good one with Lisa implying the Mona Lisa, but connotations of names shouldn’t be relevant)
However, what I was really curious about was the redwood question. Surprisingly, I knew the answer pretty much exactly and later noticed the complaints about using feet as a unit which was rather strange. I remembered this exact question was from a recent study in a psychology journal on cognitive biases.
It seems likely the question was written with knowledge of the study in mind, to test for anchoring bias, considering many other things could have been chosen other than the height of redwood trees in feet. I really hope the intention was not to take these results to say, “LW readers conclusively have less anchoring bias than people in the psychology study.” As the maximum number that could possibly be generated by the RNG (999) iirc is lower than an anchoring number given in the study (1000), presumably few survey participants if any would answer the tree height is greater than that. Due to the RNG anchor likely being much closer to the true value anchoring bias will be less than in the psychology study by nature of how the question was presented.